3 - 8 years
5 - 12 Lacs
Posted:2 days ago
Strong proficiency in Python. Experience with PyTorch and/or TensorFlow. Good understanding of ML and DL concepts. Experience with NumPy, Pandas, OpenCV, and related libraries.
- Hands-on experience building OCR pipelines and understanding of image preprocessing, text detection, text recognition, and document layout analysis. Experience with PDF processing, scanned documents, tables, forms, and structured data extraction. Familiarity with OCR evaluation and optimization techniques.- Practical experience with LLMs such as Llama, Qwen, Mistral, or similar models; understanding of Transformer architectures and attention; experience with prompt engineering and structured information extraction; experience with Vision-Language Models is desirable.- Hands-on experience with LoRA and QLoRA fine-tuning; familiarity with Hugging Face Transformers, PEFT, and TRL; experience with supervised fine-tuning; knowledge of quantization and GPU-based training/inference.- Experience building REST APIs using FastAPI or Flask; familiarity with Docker and Linux environments; understanding of model serving and production deployment; experience with GPU-based workloads and CUDA is a plus.- Experience building document intelligence or intelligent document processing (IDP) solutions; handwritten text recognition; extracting information from invoices, bank statements, identity documents, legal documents, or financial records; familiarity with layout-aware models such as LayoutLM and Donut; OCR datasets and annotation tools; knowledge of RAG and multimodal AI architectures.- Hands-on developer who has trained or fine-tuned ML/LLM models; strong problem-solving and analytical skills; ability to experiment with different OCR and LLM architectures; understanding of the ML lifecycle from data prep to production deployment.
KEY RESPONSIBILITIES
- Design, develop, and deploy end-to-end OCR solutions using Deep Learning and LLM-based architectures.
- Work with OCR frameworks and models such as Tesseract, PaddleOCR, EasyOCR, TrOCR, Donut, and similar technologies.- Integrate Vision-Language Models (VLMs) and LLMs for document understanding and information extraction.- Fine-tune pre-trained LLMs and vision-language models using LoRA, QLoRA, and other PEFT techniques.- Build document intelligence pipelines for text extraction, layout understanding, table extraction, and key-value extraction.- Process complex documents with varying layouts, low-quality scans, handwritten text, and multiple languages.- Prepare, clean, annotate, and manage datasets for model training and evaluation.- Optimize model performance, inference speed, accuracy, and GPU utilization.
Work from Office
Full Time
Net Connect
Upload Resume
Drag or click to upload
Your data is secure with us, protected by advanced encryption.
Browse through a variety of job opportunities tailored to your skills and preferences. Filter by location, experience, salary, and more to find your perfect fit.
We have sent an OTP to your contact. Please enter it below to verify.
Noida, Gurugram, Delhi
5.0 - 12.0 Lacs P.A.
Noida, Gurugram, Delhi
5.0 - 12.0 Lacs P.A.
AI Job Matching Summary